Abstract: Recently, security has been a growing concern for human life, and the cost is a significant concern. The research recommends implementing a real-time recognition system to rapidly handle photographs to reduce monitoring expenses and ensure safety in homes and offices by identifying people. The main goal of this study was to create an intelligent face recognition approach using deep learning for intelligent homes. To illustrate the study's effectiveness, the suggested model is contrasted with other cutting-edge techniques. The research introduces a tree-based deep model for facial recognition in the Cloud, which requires fewer computing resources but maintains accuracy. The input volume of the model is partitioned into multiple volumes, and trees are created for each volume according to their height and number of branches. A residual function, constructed from a convolutional layer and two non-linear functions, represents each branch. The proposed model is assessed on various publicly accessible databases and compared with the industry's best deep models for facial recognition.